Knock sensing ideas and circuit designs

From DIY contraptions to sophisticated FreeEMS-specific designs! Plus general hardware development!
Post Reply
davebmw
LQFP144 - On Top Of The Game
Posts: 331
Joined: Sun Jul 13, 2008 2:58 pm
Location: South Wales, UK

Knock sensing ideas and circuit designs

Post by davebmw »

OK its seems that there is some sound statistics behind why knock is rally a crap thing to have in an engine, I recently had a head gasket go pop because of it.

What this thread is all about is to pull together as much information as possible and some circuit ideas with a view to be incorporated in a future hardware release of FreeEMS maybe in this first one who knows.

I have several ideas based around my 2 stock Bosch sensors that i will be using to experiment with on MS2 to start with and if Fred and Sean can be sweet talked into putting a simple knock retard pin into the design of the code much like MS2 has then any design we come up with here may be useful on the first incarnation of FreeEMS.

I have started a parallel thread concerning Ion sensing also.
93'BMW 325is M50B25TU, Rebuilt 06/06, JE10.5:1, polish&port. Scorpion BB, K&N CAI, TEJ21 WBO2, '07 M3 Evo 18" 225F, 255R, EBC Kevlar, Bilstien Sprint, Polyflex. Head rebuild Oct'08, OEM+FSE FPR, MS2v3.0_DJB Custom, Extra 2.0.1
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Knock sensing ideas and circuit designs

Post by jharvey »

Some thoughts I've had about it. I think it's easier to make a no CPU knock sensor than it is to make a duel CPU EMS. Simple hardware that works with the TA card for now. Later we can add a CPU and do super mc duper knock, and ion sense.

Perhaps we should think of an semi adjustable frequency detector, then leave it up to soft to window, and keep track of sensativity? It's easy enought to build a band pass filter, followed by an energy dector, that could either go straight into the TA card, or run through a pot and opamp to make for a digi input.

I think hardware should be able to window based on the spark signal, as well as adjust expected noise floor changes based on RPM.

So should the TA inputs go like this, AN input for frequency dected and how much, Digi input for ping dected, and pehaps a bus (serial, can, what have you) for super mc duper knock. I think that can be done in three pins.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Knock sensing ideas and circuit designs

Post by Fred »

Dave, the first step would be some solid logging of the data being received. After that it can be analysed and a scheme come up with for retard. This is the normal approach anyway as you need to get the sensitivities correct etc.

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
davebmw
LQFP144 - On Top Of The Game
Posts: 331
Joined: Sun Jul 13, 2008 2:58 pm
Location: South Wales, UK

Re: Knock sensing ideas and circuit designs

Post by davebmw »

OK I have been working on my knock sensors (Bosch) and have scoped the actual Knock signal its a very sharp 2 volt spike. All the rest of the noise is either low and growly as the cam lobes slap the hydraulic lifters or much higher at about 12 to 23Khz and it seems to be the valves closing. unfortunately I cannot at the moment get hold of a NI USB DAQ. I am however thinking about hooking it up to the line in on the laptop audio.
What it does look like at the moment is the knock is quite distinct from the noise floor it should be quite easy to filter out the rest of the crap with a simple band pass filter.

Oh and I only need 1 pin I have the windowing logic sorted I am in the process of making the schematic for posting here.

give me 10 minutes and I'll post it up.
93'BMW 325is M50B25TU, Rebuilt 06/06, JE10.5:1, polish&port. Scorpion BB, K&N CAI, TEJ21 WBO2, '07 M3 Evo 18" 225F, 255R, EBC Kevlar, Bilstien Sprint, Polyflex. Head rebuild Oct'08, OEM+FSE FPR, MS2v3.0_DJB Custom, Extra 2.0.1
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Knock sensing ideas and circuit designs

Post by Fred »

What about 1 adc pin and variable sensitivity? I don't wanna waste an ADC pin, but I do like the idea of knowing how strong the knock was.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
davebmw
LQFP144 - On Top Of The Game
Posts: 331
Joined: Sun Jul 13, 2008 2:58 pm
Location: South Wales, UK

Re: Knock sensing ideas and circuit designs

Post by davebmw »

If we can have an ADC pin that would be supurb! at the moment I have Yes there is a knock logic 1 windowed by ignition events and a 120 Degree listen window.
93'BMW 325is M50B25TU, Rebuilt 06/06, JE10.5:1, polish&port. Scorpion BB, K&N CAI, TEJ21 WBO2, '07 M3 Evo 18" 225F, 255R, EBC Kevlar, Bilstien Sprint, Polyflex. Head rebuild Oct'08, OEM+FSE FPR, MS2v3.0_DJB Custom, Extra 2.0.1
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Knock sensing ideas and circuit designs

Post by Fred »

davebmw wrote:If we can have an ADC pin that would be supurb! at the moment I have Yes there is a knock logic 1 windowed by ignition events and a 120 Degree listen window.
Well, about half of the 16 are spare - sparish. I think it makes sense. It will be optional anyhow.

As for 120 degrees, that seems very wide. Given that you NEVER advance ANY engine more than about 60 and it CAN'T knock much after TDC I'm pretty sure it should be narrower. Something like 20 btdc and 10atdc or so ?

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
davebmw
LQFP144 - On Top Of The Game
Posts: 331
Joined: Sun Jul 13, 2008 2:58 pm
Location: South Wales, UK

Re: Knock sensing ideas and circuit designs

Post by davebmw »

OK here is the crude ignition event based knock detector:
Printing KnockSensing-sch.pdf
(14.99 KiB) Downloaded 1817 times
Basic principle of operation is that the front end comparator is set using a preset to trigger at a certain level about 1.2V should do it.
The 10nF decoupling caps de-sensitise the comp from lower end noise.

the 6 ignition signals come straight from the gate drives for the ignition drivers so they get a logic 1 very briefly so this is latched into the first pair of RS flip flops this pair provide the D signal to the second pair that are clocked every time the threshold is attained on the 2 comparators.
the circuit is very crude but it listens to the right sensor at the right time and captures the signal based on ignition events.

With the addition of a bandpass filter on the front end it could be very effective.

An engine will knock after TDC if the conditions dictate, until it has passed the peak pressure point about 30 to 40 degrees down the power stroke you can't really say you are out of the danger zone.

Anyway its using simple discrete logic if I had an output from the XDP that gave me a pulse per degree I could get it a lot more selective :p

When you get 5 read that paper I put up in ion sensing its cooool!
93'BMW 325is M50B25TU, Rebuilt 06/06, JE10.5:1, polish&port. Scorpion BB, K&N CAI, TEJ21 WBO2, '07 M3 Evo 18" 225F, 255R, EBC Kevlar, Bilstien Sprint, Polyflex. Head rebuild Oct'08, OEM+FSE FPR, MS2v3.0_DJB Custom, Extra 2.0.1
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Knock sensing ideas and circuit designs

Post by Fred »

actually, the ign pulses aren't short relative to the total time at higher revs.

17ms at 7500 and dwell per coil of 5ms = long pulse that varies its start timing with rpm.

you really want to trigger off the falling edge of it and with a very short pulse of some sort. I have no idea about this stuff, but you do :-)

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
davebmw
LQFP144 - On Top Of The Game
Posts: 331
Joined: Sun Jul 13, 2008 2:58 pm
Location: South Wales, UK

Re: Knock sensing ideas and circuit designs

Post by davebmw »

Ok thats should just be as simple as adding an inverter in the IG1 to 6 coil drive lines, as the flip flops clock on the rising edges only.
93'BMW 325is M50B25TU, Rebuilt 06/06, JE10.5:1, polish&port. Scorpion BB, K&N CAI, TEJ21 WBO2, '07 M3 Evo 18" 225F, 255R, EBC Kevlar, Bilstien Sprint, Polyflex. Head rebuild Oct'08, OEM+FSE FPR, MS2v3.0_DJB Custom, Extra 2.0.1
Post Reply